Commissioning of respiratory-gated 4D dynamic dose calculations for various gating widths without spot timestamp in proton pencil beam scanning.

Background: Proton pencil beam scanning (PBS) is susceptible to dose degradation because of interplay effects on moving targets. For cases of unacceptable motion, respiratory-gated (RG) irradiation is an effective alternative to free breathing (FB) irradiation. However, the introduction of RG irradiation with larger gate widths (GW) is hindered by interplay effects, which are analogous to those observed with FB irradiation. Accurate estimation of interplay effects can be performed by recording spot timestamps. However, our machine lacks this feature, making it imperative to find an alternative approach. Thus, we developed an RG 4-dimensional dynamic dose (RG-4DDD) system without spot timestamps.

Purpose: This study aimed to investigate the accuracy of calculated doses from the RG-4DDD system for PBS plans with varying breathing curves, amplitudes, and periods for 10%-50% GW.

Methods: RG-4DDDs were reconstructed using in-house developed software that assigned timestamps to individual spots, integrated start times for spills with breathing curves, and utilized deformable registrations for dose accumulation. Three cubic verification plans were created using a heterogeneous phantom. Additionally, typical liver and lung cases were employed for patient plan validation. Single- and multi-field-optimized (SFO and IMPT) plans (ten beams in total) were created for the liver and lung cases in a homogeneous phantom. Lateral profile measurements were obtained under both motion and no-motion conditions using a 2D ionization chamber array (2D-array) and EBT3 Gafchromic films on the CIRS dynamic platform. Breathing curves from the cubic plans were used to assess nine patterns of sine curves, with amplitudes of 5.0-10.0 mm (10.0-20.0 mm target motions) and periods of 3-6 sec. Patient field verifications were conducted using a representative patient curve with an average amplitude of 6.4 mm and period of 3.2 sec. Additional simulations were performed assuming a ± 10% change in assigned timestamps for the dose rate (DR), spot spill (0.08-s), and gate time delay (0.1-s) to evaluate the effect of parameter selection on our 4DDD models. The 4DDDs were compared with measured values using the 2D gamma index and absolute doses over that required for dosing 95% of the target.

Results: The 2D-array measurements showed that average gamma scores for the reference (no motion) and 4DDD plans for all GWs were at least 99.9 ± 0.2% and 98.2 ± 2.4% at 3%/3 mm, respectively. The gamma scores of the 4DDDs in film measurements exceeded 95.4% and 92.9% at 2%/2 mm for the cubic and patient plans, respectively. The 4DDD calculations were acceptable under DR changes of ±10% and both spill and gate time delays of ±0.18 sec. For the 4DDD plan using all GWs for all measurement points, the absolute point differences for all validation plans were within ±5.0% for 99.1% of the points.

Conclusions: The RG-4DDD calculations (less than 50% GW) of the heterogeneous and actual patient plans showed good agreement with measurements for various breathing curves in the amplitudes and periods described above. The proposed system allows us to evaluate actual RG irradiation without requiring the ability to record spot timestamps.